First store with populated D41 claim fields (post-#151, eb480b7, luna bindings): 37/410 claims carry structured valid-time, including Caroline went to an LGBTQ support group on 2023-05-07 → [2023-05-07..2023-05-07, day, event_time].
Yet entity_timeline for the main entity still returns a single bucket: {"2026": 133} — ingestion year. The recipe aggregates fact-layer validity (relations/observations), and nothing carries claim_valid_* from claims through E3/normalization into the fact layer the timeline queries. #151 populated the ground floor; the surfaces that motivated it (entity_timeline, as-of reads, WP-8.5) still can't see it.
Needs a design decision on propagation semantics (a fact's valid-time versus its supporting claims' valid-times is a real D41/D43 question, not a mechanical copy), then the wiring.
🤖 Generated with Claude Code
https://claude.ai/code/session_01GKENhTLJg1HqhbdwCmmkbc
First store with populated D41 claim fields (post-#151, eb480b7, luna bindings): 37/410 claims carry structured valid-time, including
Caroline went to an LGBTQ support group on 2023-05-07→[2023-05-07..2023-05-07, day, event_time].Yet
entity_timelinefor the main entity still returns a single bucket:{"2026": 133}— ingestion year. The recipe aggregates fact-layer validity (relations/observations), and nothing carriesclaim_valid_*from claims through E3/normalization into the fact layer the timeline queries. #151 populated the ground floor; the surfaces that motivated it (entity_timeline, as-of reads, WP-8.5) still can't see it.Needs a design decision on propagation semantics (a fact's valid-time versus its supporting claims' valid-times is a real D41/D43 question, not a mechanical copy), then the wiring.
🤖 Generated with Claude Code
https://claude.ai/code/session_01GKENhTLJg1HqhbdwCmmkbc